AI Safety's Biggest Talent Gap Isn't Researchers. It's Generalists.

A new analysis from AI safety leaders at Kairos and Constellation reveals a severe and growing talent imbalance in the field. While over 20 research fellowships are on track to produce 2,000-2,500 technical and policy researchers this year, the pipeline for non-research talent is critically underdeveloped. Only about 7 fellowships exist for generalists—roles like program managers, operators, fieldbuilders, and founders—collectively producing roughly 300 fellows. Senior leaders report that non-research job postings often attract hundreds of applications but yield only 0-5 candidates with the necessary mission alignment, AI safety context, and core competence, creating a major bottleneck for organizational growth and project execution.

In response, Kairos and Constellation are announcing the Generator Residency, a new program designed to build a legitimate career ladder for generalists. The initiative aims to fix a system that currently incentivizes ambitious junior talent to pursue legible technical fellowships, even if their skills are better suited for operational or leadership roles. The residency seeks to make generalist careers—encompassing everything from project management and communications to organizational leadership and cybersecurity—as accessible and respected as the well-worn research path from programs like ARENA, SPAR, and MATS. This is a direct effort to cultivate the people who can manage research, run organizations, and implement safety ideas at scale.